The Role of Disturbances for the Antarctic Benthos
- A Simulation Study -
an der Fakultät für Mathematik und Naturwissenschaften der
C
ARL VONO
SSIETZKYU
NIVERSITÄTO
LDENBURG zur Erlangung des Grades einesDoktors der Naturwissenschaften (Dr. rer. nat.)
vorgelegte Dissertation
von Diplom Biologe Michael Potthoff
Gutachter: Privatdozent Dr. habil. Julian Gutt Zweitgutacher: Prof. Dr. Pedro Martinez Arbizu
I REMEMBER MY FRIEND JOHNNY VON NEUMANN USED TO SAY,“WITH FOUR PARAMETERS I
CAN FIT AN ELEPHANT AND WITH FIVE I CAN MAKE HIM WIGGLE HIS TRUNK.”
Enrico Fermi, quoted by Freeman Dyson in “A Meeting with Enrico Fermi” (Nature 427 p297.)
Aufgrund ihrer Unzugänglichkeit sind marine Systeme, insbesondere die benthischen Gemeinschaften des antarktischen Schelfs, schwer zu untersuchen.
Simulationsmodelle sind ein alternatives Werkzeug, um Hypothesen und Ideen über relevante Prozesse und Mechanismen dieser Systeme zu untersuchen und zu testen. Mit unterschiedlichen Simulationsmodellen wurden die Bedeutung und der Einfluss verschiedener Faktoren, wie Ausbreitungsmuster und Lebensspanne einzelner Arten und Störungen durch Eisbergstrandungen für antarktische Schelfgemeinschaften untersucht.
Mit einem Simulationsmodell wurde die Bedeutung eines fleckenartigen Larvenausbreitungs- und Rekrutierungsmusters untersucht. Solche Muster sind in marinen Lebensräumen häufig anzutreffen. Mit dem Modell konnte gezeigt werden, dass diese Muster eine hinreichende Bedingung für Artenkoexistenz darstellen und somit förderlich für die Diversität sind.
Generell wird für viele antarktische Arten ein eingeschränktes Ausbreitungspotential angenommen. Als Grund hierfür wird die starke Saisonalität der Primärproduktion in hohen Breiten gesehen. Anhand eines Simulationsmodells konnte gezeigt werden, dass in einem dynamischen System, in dem neue Lebensräume durch Störungen entstehen, aufgrund der Langlebigkeit der Individuen eine weit reichende Ausbreitung für das lokale Überleben einer Pionierpopulation nicht notwendig ist.
Generell hat die Langlebigkeit für die Ausbreitungsdistanz eine große Bedeutung.
Die minimale Ausbreitungsdistanz, die zum Überleben einer Pionierpopulation notwendig ist, hat eine nicht-lineare, hyperbolische Abhängigkeit zur Lebensdauer.
Daher profitieren besonders kurzlebige Arten von einer möglichen Verlängerung ihrer Lebensspanne. Eine Art könnte mit der halben Ausbreitungsdistanz auskommen, wenn sie ihr Überleben um das drei bis vierfache verlängern könnte.
Basierend auf Daten über den Störungsumfang und die Lebensdauer von Pionierorganismen aus dem Weddell-Meer könnte ein sechsjähriger Primärbesiedler von Eisbergkratzerspuren mit einer Ausbreitungsdistanz von weniger als 1000 m auskommen. Daher kann das eingeschränkte Ausbreitungspotential vieler antarktischer Arten eine Anpassung an den hoch dynamischen Lebensraum darstellen.
In weiteren Simulationen wurde der Einfuß verschiedener physikalischer Eigenschaften einer Störung, wie Störungsgröße und Frequenz, auf die Sukzession untersucht. Die Simulation zeigte, dass es einen Unterschied zwischen kleinen, zahlreichen und wenigen, aber großen Störungen bei gleicher Gesamtgröße des Störungsareals gibt. Dabei bestimmt das Ausbreitungspotential der vorhandenen Arten die Reaktion und Zusammensetzung der entstehenden Artengemeinschaft.
Weiterhin zeigte sich ein großer Einfluss von Überlebenden einer Störung auf die Elastizität der Gemeinschaft. Schon eine Überlebenswahrscheinlichkeit von 1% kann die Regenerationsdauer um bis zu 25% verkürzen. Dieser Effekt ist auf einen so genannten „räumlichen Speicher-Effekt“ zurückzuführen. Ähnlich wie z.B. Bäume, die nach einem Waldbrand wieder ausschlagen und fruchten, wird besonders bei Arten mit geringem Ausbreitungsvermögen die Zeit zur Besiedlung gestörter Gebiete stark reduziert.
Schätzungen für die direkte Mortalität durch Eisbergstrandungen für das Benthos liegen nahe bei 99%. Allerdings könnten durch unregelmäßige Kielform von Eisbergen Bereiche zwischen gestörten Flächen unberührt bleiben. Dadurch entsteht ein Mosaik aus verschieden stark gestörten Flächen, die sich ähnlich positiv auf die Elastizität antarktischer Gemeinschaften auswirken.
Due to their remoteness marine systems in general and the Antarctic benthic shelf assemblages in special are difficult to investigate. Computer simulation models provide an alternative tool to test hypotheses and ideas on the process that structure and influence these systems. Several computer simulation models were used to explore the importance of different factors like dispersal pattern and longevity of single species and disturbance due to iceberg scouring for benthic Antarctic shelf assemblages.
In a simulation the outcome of a clumped or patchy larvae dispersal and settlement pattern was analysed. Such patterns are commonly found in marine systems. In the simulation they were found to be a sufficient condition for species coexistence and thus enhanced diversity.
It is often assumed that Antarctic species have a limited dispersal potential. The reason thereof is seen in the high seasonality of primary production at high latitudes.
A simulation model revealed that in a dynamic environment, where suitable habitat is the result of disturbances, the species longevity can make long range dispersal unnecessary for the local persistence of a pioneer population. The longevity plays a central role for the minimal dispersal distance. This minimal dispersal distance has a non-linear dependency on species longevity. Thus especially short living species can profit much from a prolonged lifespan. A species can cope with a halved dispersal distance if it could extend its lifespan three to four times.
Based on disturbance and pioneer lifetime data from the Weddell Sea a minimum dispersal distance of less than 1000 m might be sufficient for a primary coloniser of iceberg scours with a lifespan of about 6 years. Thus the limited dispersal potential of Antarctic species can be an adoption the highly dynamic environment.
In a further simulation the role of physical disturbance properties such as disturbance size and frequency was explored. The simulation showed a difference between numerous, small and rare large disturbances with the same total perturbed area. The dispersal limitation of the involved species influenced the community structure and response to different disturbance regimes. Additionally single surviving individuals had great influence on the resilience of the assemblages. Even a 1% survival probability reduced the recovery time up to 25%. This effect can be attributed to a spatial storage effect. Similar to trees that survived a burning and re-sprout, the time
to invade recently disturbed habitats can be dramatically reduced, especially for species with a high dispersal limitation.
Estimates on the severity of iceberg scouring to the Antarctic benthos are close to 99%. However, irregular keel forms may lead to undisturbed areas between scour marks. By this a mosaic of areas in different states emerges, that has a similar positive effect for the resilience of Antarctic assemblages.
1. THE ANTARCTIC ... 1
1.1 THE ANTARCTIC, ORIGIN AND CURRENT CONDITIONS... 2
1.2 THE ANTARCTIC ENVIRONMENT... 3
1.3 ANTARCTIC BENTHIC MARINE LIFE... 5
1.4 REPRODUCTION AND DISPERSAL TRAITS OF ANTARCTIC SPECIES... 5
1.5 THE INFLUENCE OF ICE... 7
1.6 ICEBERGS... 8
1.7 ICEBERG POPULATION... 9
1.8 ICEBERG DRIFT... 10
1.9 INFLUENCE OF ICEBERGS ON THE ANTARCTIC SEA FLOOR... 14
2. ECOLOGICAL MODELLING ... 15
2.1 CONCEPTIONAL MODELS... 15
2.2 DESCRIPTIVE MODELS... 16
2.3 SIMULATION MODELS... 16
2.3.1 Spatial explicit models ... 19
2.3.2 Individual Based Models ... 19
2.3.3 Simulation Models and Biodiversity ... 19
2.4 THE MODELLING CYCLE... 20
3. THE AIM OF THIS STUDY... 21
4. PUBLICATIONS AND MANUSCRIPTS ... 24
4.1 “CLUMPED DISPERSAL AND SPECIES COEXISTENCE” ... 24
4.1.1 Bibliographic record ... 24
4.1.2 Short summary of ideas, problems, solutions and results:... 24
4.2 “HOW TO SURVIVE AS A PIONEER SPECIES IN THE ANTARCTIC BENTHOS: MINIMUM DISPERSAL DISTANCE AS A FUNCTION OF LIFETIME AND DISTURBANCE” ... 25
4.2.1 Bibliographic record ... 25
4.2.2 Short summary of ideas, problems, solution and results: ... 25
4.3 “HOW THE DISTURBANCE SEVERITY DRIVES THE BENTHIC DIVERSITY ON THE ANTARCTIC SHELF” ... 27
4.3.1 Bibliographic record ... 27
4.3.2 Short summary of ideas, problems, solution and results: ... 27
5. SUMMARY AND DISCUSSION ... 29
6. LITERATURE ... 34
7. CLUMPED DISPERSAL AND SPECIES COEXISTENCE... 39
7.1 ABSTRACT... 40
7.2 INTRODUCTION... 41
7.3 THE MODEL... 44
7.3.1 Homogeneous versus heterogeneous environments ... 44
7.3.2 Species definition... 44
7.3.3 Local dispersal ... 45
7.3.4 Patchy dispersal ... 45
7.3.5 Local dynamics... 46
7.3.6 Constant versus fluctuating environments ... 46
7.4 COMPUTER SIMULATION EXPERIMENTS... 46
7.5 RESULTS... 48
7.5.1 Experiment 1 (neutral model)... 49
7.5.2 Experiment 2 (different mean dispersal distances, single dispersal mode) ... 49
7.5.3 Experiment 3 (different mean dispersal distances and two dispersal modes) . 49 7.6 DISCUSSION... 52
7.7 LITERATURE CITED: ... 56
8. HOW TO SURVIVE AS A PIONEER SPECIES IN THE ANTARCTIC BENTHOS... 59
8.1 ABSTRACT... 60
8.2 INTRODUCTION... 61
8.3 METHODS... 62
8.4 COMPUTER SIMULATIONS... 63
8.4.1 Habitat spacing ... 63
8.5 MINIMUM DISPERSAL DISTANCE OF A PIONEER SPECIES (DMIN) ... 64
8.6 RESULTS... 65
8.6.1 Habitat spacing ... 65
8.6.2 Minimum dispersal distance... 66
8.6.3 Dependency of the minimum dispersal distance on longevity... 67
8.7 APPLICATION OF THE RESULTS TO THE BENTHIC ASSEMBLAGES OF THE WEDDELL SEA 69 8.8 DISCUSSION... 70
8.8.1 The Simulation Model ... 70
8.8.2 Relevance for the Antarctic communities... 72
8.9 APPENDIX... 75
8.10 BIBLIOGRAPHY... 76
9. HOW THE DISTURBANCE SEVERITY DRIVES THE BENTHIC DIVERSITY ON THE ANTARCTIC SHELF ... 79
9.1 ABSTRACT... 79
9.2 INTRODUCTION... 80
9.3 MATERIAL AND METHODS... 81
9.3.1 General description of the model:... 81
9.3.2 Simulation of Iceberg Scouring, Disturbance Regime ... 82
9.3.3 Species Traits ... 83
9.3.4 The Nullmodel ... 84
9.3.5 Larval dispersal... 84
9.3.6 Competition ... 85
9.3.7 Succession pattern / succession state definition... 86
9.3.8 Succession speed ... 86
9.3.9 Starting and stop conditions/ runtime ... 87
9.4 EXPERIMENTS: ... 87
9.5 RESULTS... 88
9.5.1 Experiment 1 ... 89
9.5.2 Experiment 2 ... 90
9.6 DISCUSSION... 90
9.7 APPENDIX ... 98
9.7.1 Calculation of instantaneous mortality on mean life time ... 98
9.7.2 Calculation of the rotation period... 98
9.7.3 The concept of intra-guild neutrality ... 98
TABLES... 99
10.1 INTRODUCTION... 106
10.1.1 Time in the model ... 106
10.1.2 The environment (a-biotic parameters)... 106
10.1.3 The space... 106
10.1.4 Disturbance events ... 108
10.1.5 Disturbance severity... 110
10.1.6 The flow ... 110
10.2 THE BIOLOGY MODEL OF SIMBAA(BIOTIC PARAMETER) ... 112
10.2.1 List of species traits available in SIMBAA: ... 112
10.2.2 General information ... 113
10.2.3 Reproductive Traits ... 113
10.3 DISPERSAL TRAITS... 115
10.4 THE DISPERSAL IN SIMBAA ... 116
10.4.1 Overview of dispersal kernels available in SIMBAA: ... 118
10.4.2 Lifespan / mortality ... 122
10.4.3 Other species traits... 124
10.4.4 Growth mode ... 126
10.5 THE SIMBAAGRAPHICAL USER INTERFACE (SIMBAA-GUI) ... 127
10.5.1 The main window ... 128
10.5.2 The disturbance editor ... 129
10.5.3 The species editor... 130
10.5.4 The landscape editor ... 132
10.5.5 The flow editor ... 135
10.5.6 The visualise tool window ... 136
10.5.7 The virtual ROV ... 139
10.5.8 Cluster analysis tool... 141
10.5.9 Rank analysis tool ... 144
10.5.10 Succession analysis tool, succession state definition ... 145
10.5.11 Age structure analysis tool ... 147
10.6 ADDITIONAL SIMULATION PARAMETER DIALOG... 148
10.6.1 Simulation stop conditions: ... 150
10.7 DIVERSITY MEASUREMENTS AVAILABLE IN SIMBAA... 154
10.7.1 α-Diversity ... 154
10.7.2 β-Diversity... 155
10.8 THE M-INDEX... 156
10.8.1 M-Index, an example:... 157
10.9 GENERAL SIMBAA TIPS... 162
10.9.1 Appendix... 164
10.9.2 SGF-Header ... 165
10.9.3 State definitions ... 167
10.9.4 Flow Grid definition... 168
10.9.5 Disturbance definitions ... 169
10.9.6 Species definitions ... 170
10.9.7 Simulation grid definition... 171
10.10 ACKNOWLEDGEMENT, EXTERNAL CODE... 173
1. The Antarctic
Figure 1, Distribution of the Sub-Antarctic and Polar Fronts and associated currents in the Antarctic. The approximate positions of the Weddell Gyre and the Ross Sea Gyre are also shown. The Antarctic Divergence is between the Polar Current and the Antarctic Circumpolar Current. Blue-grey shadings indicate water depth less than 3000 m (modified after Brown et al. 1990).
Error! Style not defined.
2
1.1 The Antarctic, origin and current conditions
The Antarctic (Figure 1) includes the ice covered continental landmass encompassing the South Pole and the surrounding southern ocean. The landmass originates from the super-continent Gondwana, which started to break up around the late Jurassic (170 mya). At this time the isolation of the Antarctic started. During Eozaen, 55 mya ago, the separation and north drift of Australia occurred. About 26 mya before present, the opening of the Drake Passage cut off the last remaining land bridge to South America (Walter 2005). Since the final establishing of the deep sea trench forming the Drake Passage, the Antarctic continent has no direct connection to other landmasses. This facilitated the formation of a ring ocean around Antarctica. Several wind systems drive the current system of this ocean. The main components are two counterwise rotating currents. The outer, most northward one is the Westwind drift with a speed of about 0.5 km h-1. Due to the lack of a land barrier, it floats clockwise around the entire continent and is termed Antarctic Circumpolar Current (ACC). The ACC affects the water body down to the bottom of the sea (Walter 2005). It consists of several water fronts. Its northern most border is the Sub- Antarctic Front. The Polar Front (referred to as the Antarctic Convergence in older literature) is located around 50 degree south, but its exact location changes with the seasons (Figure 1). The Polar Front separates the cold Antarctic waters from warmer oceans. With a temperature gradient, the water temperature falls from about 10°-8°C to a temperature just below 2°C, it is a strong physiological barrier limiting the faunal exchange with other biota.
The Antarctic divergence is located around the 65th degree south. South of this zone strong eastward winds cause an anti-clockwise rotating current system. Due to the Corilois force, originating from the earth rotation, this current receives a southward impulse and is pushed against the continent. Over the shelf region this results in a strong, westward running current parallel to the cost, termed Antarctic costal current, Polar current or simply Eastwind Drift. This steady and homogeneous current affects the environmental conditions over the Antarctic shelf region. The Antarctic shelf is generally deeper than other continental shelfs. Due to the ice shields on the continental landmass that pushes down the continental plate it can reach up to 800 m in depth.
1.2 The Antarctic environment
The Antarctic marine environment is characterised by several abiotic factors (see Arntz et al. 1994). The constant low water temperature shows only a small annual amplitude and the fluctuations in seawater salinity are also low. Contrarily the sea ice cover exhibits strong seasonal fluctuations. During the austral winter it covers up to two-thirds of the southern ocean (~20*106 km²) and strongly restricts the light irradiation depth in the water body. In this period only icealgae within a lacuna- system in the sea ice and, to some extend, in the ice-water contact zone may be photosynthetic active. Primary production in the open water is restricted to the short period of the austral summer when the sea ice has retreated to about 4*106 km².
When sea ice bakes up in spring, the release of ice algae from their confined habitats and leads to a first spring bloom (Bathmann 1995). A second algae bloom follows this spring bloom later in summer (Bathmann 1995, Lochte and Smetacek 1995).
Consequently, organic matter supply to the sea floor is mainly restricted to the summer period. At some locations this seasonal food rain accumulates to several centimetre thick layers on the sea floor (Figure 2). Such a situation may be comparable to the North Atlantic, where thick algae mats, originating from under-ice algae, rapidly sink to the abyssal sea floor and present strong local food pulses in spring (Schewe and Soltwedel 2003).
Error! Style not defined.
4
Figure 2: An approximately 3 cm thick phytodetritus layer on the sea floor of the Antarctic shelf (Lazarev Sea). In the upper left part a holothurian is visible, with its body almost completely immersed in the phytodetritus. According to Mincks et al. (2005) such detritus layers can persist for years and serve as benthic food banks. We postulate that iceberg scouring may be a major source of resuspension that makes the phytodetritus available to benthic filter feeders, especially in winter. (Photo © J.Gutt / AWI)
From a coarser ecological view the environmental conditions over the Antarctic shelf resemble a relatively homogeneous ring around the continent, mostly with a width of less than 100 km. At some locations huge parts additionally are below a permanent shelf ice cover. The rich marine life and these steady conditions led Dayton et al.
(1994) to realise “Perhaps the most interesting question of polar marine biology relates to the fact that Antarctic has a much higher species richness (…) yet lacks the ecological diversity of the Arctic”.
1.3 Antarctic Benthic Marine Life
The diversity of the shelf fauna, especially of the sedentary species, is high. Gutt et al. (2004) estimated a total of around 17.000 species. Many sedentary species belong to the epifauna, living on the sediments of the shelf or colonise biogenic structures (e.g. sponges). Compared to other soft sediment systems (e.g. the North Sea or Deep Sea), relatively few infauna elements exist. The comparable high degree of endemism (in some groups, e.g. amphipods 80%, pycnogonids 90% see Clarke and Johnston 2003) as well as the dominance of some groups, e.g.
hexactinellid sponges (Barthel 1992) is conspicuous, as is the deficit in other groups like higher decapods or chondrichtheys. This fauna inventory is a direct consequence of the evolutional and thus geo- and glacial history of Antarctica (Thatje et al. 2005).
One important factor for the recent fauna composition is the isolation of the continent, restricting the faunal exchange with other biota and fostering the radiation of successful colonisers.
1.4 Reproduction and dispersal traits of Antarctic species
The Antarctic life is considered to be strongly influenced by the conditions of the Antarctic, representing a permanent cold but highly seasonal environment. The proposed effects include slow, seasonal growth, prolonged lifespan, low mortality and large adult size (Arntz et al. 1994). Regarding the reproductive traits, Arntz et al.
(1994) listed a prolonged gametogenesis and a delayed maturation. They further proposed a general low fecundity, slow embryonic development and a seasonal reproduction pattern to be typical for Antarctic species. In general Antarctic species seem to produce either large yolky eggs with a non-pelagic development or show brood protection by brooding or viviparity (Arntz et al. 1994). Besides the slowing down of biological processes due to the cold, one reasons for such life history traits are seen in the strong seasonality (Clarke 1990, Brockington and Clarke 2001).
A general lack of meroplanktic development was proposed for the Antarctic since the first early expeditions (see Pearse et al. 1991, Arntz et al. 1994). The apparent decline in planktonic larvae from low to high latitudes led to the formulation of
“Thorson’s rule” (Mileikovsky 1971). This theory was originally based on Thorson’s analysis of prosobranch gastropod larvae (see Gallardo and Penchaszadeh 2001).
After Thorson’s concept the amount of species with planktotrophic larvae and an
Error! Style not defined.
6
indirect development should decrease from the equator towards the poles whereas the number of species with lecitotrophic larvae and direct development rises. The main reason for such a shift is seen in the increasing seasonality of the primary production towards high latitudes. Low or nearly absent numbers of meroplanktic larvae in plankton samples seem to be in line with this hypothesis as well as the high amount of species with brooding or budding behaviour in Antarctic assemblages.
However, the number of papers dealing with meroplanktic larvae in south polar waters continuously rose in the past years (see Shreeve and Peck 1995, Bhaud et al.
1999, Stanwell-Smith et al. 1999, Absher et al. 2003, Freire et al. 2006 and Sewell 2005) and Thorson’s rule is controversially discussed. Nevertheless, the general larval density within the water body as well as the known number of larval morphotypes seems to be rather low when compared to other regions (Thatje et al.
2005).
Thatje et al. (2005) proposed that the glacial history of the Antarctic might be a reason for a disadvantage of meroplanktic development in the southern ocean. The current disturbance regime may also explain a part of the riddle.
Recently David Bowden published the results of a comprehensive three-year study on recruitment and settling of sedentary species on artificial settling plates at Ryder Bay, Antarctic Peninsula (Bowden 2005b, 2005a, Bowden et al. 2006). This work resumes the currently available data on settling experiments in Antarctic waters. The experiments were conducted in shallow, near shore waters, but the general findings may be representative and confirm some of the hypothesis of Arntz et al. (1994):
• Colonisation and growth speed is low
• Recruitment generally can occur throughout the year. However, recruitment of single taxa seems to be strongly seasonal with an overall peak in later winter
• Growth of most species seems to be highly seasonal and coincident with the period of primary production
• Some species recruit throughout the year (e.g. spirorbids polycheates) or show exceptional fast growth (e.g. Ascidia sp., see also Rauschert (1991) and its discussion by Bowden).
• Assemblage composition seems to be controlled by post-settlement mortality (predation and disturbance by ice)
1.5 The influence of ice
After the formation of the Antarctic Circumpolar Current a period of cooling down followed, resulting in the present climate situation (see Clarke 1990, Thatje et al.
2005). According to recent data (Holbourn et al. 2005), the final cool down and the formation of the Antarctic ice shields happened within a relatively short time span about 14 mya bp. Most probably a cyclic expansion and reduction of the Antarctic ice in terms of longer time periods has occurred since then. During glacial maxima, the shelf ice eventually covered the complete shelf (Anderson et al. 2002). Based on deep scour marks on the shelf, it is evident that the ice grounded at least at many locations (see Thatje et al. 2005). For the Weddell Sea it must be assumed that the ice grounding line was along the current 500 m depth line 13.000-27.000 years ago (Anderson et al. 2002). Therefore, these areas were not available for colonisation to any species.
There is evidence of a geographical different glacial history for the east and west Antarctic (see Thatje et al. 2005). It is possible, that the fauna moved lateral along the continental shelf to escape the advancing ice. It is also possible that some refuge areas persisted, e.g. deep trenches or pockets under the ice sheet that enabled some species to survive. An other common hypothesis assumes that a majority of the fauna migrated to the continental slopes and recolonised the shelf after the retreat of the ice. The general eurybathy, the wide depth range of many Antarctic species, is often explained this way (Brey et al. 1996). Recently Thatje et al. (2005) supposed that the migration to the slopes would have exposed any species to severe disturbance due to a high amount of suspended sediments and turbidity flows caused by the advancing glaciers. Thus, the slopes must be considered as an unsuitable habitat to survive a glacial period.
Additionally, a much higher sea ice cover must be expected during a glacial, resulting in a severe reduction of the primary production (Bonn et al. 1998) with according consequences for all trophic levels. Thus, the glacial history of the Antarctic is a sequence of unstable, changing environmental conditions. However, the changes occurred probably not rapidly but over relative long time scales (hundreds or even some thousand years). Species can cope with such gradual climate changes when some refuge possibilities, e.g. alternative habitats, exist or when they are pre- adopted.
Error! Style not defined.
8
1.6 Icebergs
Regarding shorter time scales, it is known that icebergs are the major source for disturbance to the shelf fauna (Gutt et al. 1996, Gutt 2001). In the Antarctic a typical iceberg calves form the shelf ice. This results in a shape with strait edges and a flat top (see Figure 3). Therefore, Antarctic icebergs are often characterised as tabular icebergs. Due to their regular shape the draught is roughly 7 times the height above the sea surface (which can be up to 100 m according to Wüthrich and Thannheiser).
However, as icebergs become older and disintegrate, they might loose their regular shape. The thickness of a tabular iceberg is determined by the thickness of the shelf ice it originates from. The shelf ice thickness, and thus iceberg thickness, varies between 150 m and 550 m with a mean around 250 m (Gladstone et al. 2001, see Figure 4).
1
7
1
7
a b
7b 3a
Figure 3, Comparison and characteristics for Arctic (left) and Antarctic (right) icebergs. Roughly 1/8th of the volume is visible as the “tip of an iceberg”. Due to their origin from glaciers Arctic icebergs tend to have an irregular shape and extending sideways under the ocean surface. This makes them dangerous for shipping. The draught can be estimated to be 3 times the visible height (a). Antarctic icebergs normally have a more regular shape with a flat top and do not extend much sideways.
Due to their regular shape, the draught is approximately 7 times the height above the water line (b). (Changed and redrawn according to Wüthrich and Thannheiser).
<250m
Iceberg Size Distribution (after Gladstone et al. 2001)
0 0.05 0.1 0.15 0.2 0.25 0.3
40x60
67x100
133x200
175x350
333x500
467x700
600x900
800x1200
1067x1600
1467x2200 iceberg size class
percentage
0 0.25 0.5 0.75 1
cum.percentage
cum.percentage
Figure 4, Size class distribution of calving icebergs according to Gladstone et al. (2001). Size classes (x-y dimensions) smaller than 175x350 m are assumed to have a draught <250 m (left panel). These make up to ~75% of all icebergs. Bigger icebergs with a draught >= 250 m (right panel) are rare (~25%).
1.7 Iceberg population
Icebergs are possible shipping dangers. Therefore, data on shipboard iceberg sightings and size classes have been recorded since many years (Hamley and Budo 1986). Estimates for the population of icebergs south of the Antarctic convergence are in the order of 2*105-3*105 (Orheim 1987, 1988). However, this data may be biased (Gladstone and Bigg 2002) and satellite remote sensing has become a more accurate tool for tracking of iceberg population and drift today (Young et al. 1998, Gladstone and Bigg 2002, Silva and Bigg 2005). Satellite observations, however, only cover a limited area. A true census of the Antarctic iceberg population using satellite images has not been carried out yet.
Error! Style not defined.
10
1.8 Iceberg drift
Several forces influence the drift of an iceberg. Main components are the water and air drag forces. Thus icebergs move within a water body. The east wind drift takes the icebergs on a track along the coast. Strong katabatic winds may further accelerate an iceberg and cause a drift speed higher than the surrounding water (Gladstone et al. 2001). Therefore the average drift speed for the Weddell Sea is estimated to be 7.5 km d-1 near the coast and only 3.5-4.5 km d-1 for the open ocean (Gladstone and Bigg 2002). Older data (Tchernia and Jeannin 1984) reported a slightly higher speed (10.4 km d-1), with strong variation (1.8-55 km d-1). The same authors point out that iceberg drift occurs the whole year round; hence icebergs are not stopped by the winter sea ice. Contrarily, Lichey and Hellmer (2001) estimate that sea-ice strongly influences the drift and a sea-ice cover over 90 % may lock icebergs.
Thus, the observation of Tchernia and Jeannin (1984) must be explained by the fact that they used radio satellite beacons and no visual observation methods. When their tagged icebergs got locked within the sea ice, they simply drifted together with it.
Nevertheless, this points out that iceberg drift can impact the ecosystem throughout the whole year.
Figure 5, Iceberg drift pattern according to the model of Gladstone et al.
(2001). Solid arrows indicate dominant drift directions; dashed line shows the northern border of the iceberg concentration near the coastline. Colours represent the depth scale (redrawn and modified from fig.3 in Gladstone et al.
2001).
If an iceberg enters the circum polar current, it may travel around the whole Antarctic within a few years. However, most icebergs stay close to the coast over the shelf area (Figure 5, see Gladstone et al. 2001, Gladstone and Bigg 2002, Silva and Bigg 2005). Gladstone and Bigg (2002) reported a 20-35 times higher iceberg concentration near the coast than further north during a satellite based study of iceberg drift in a 400 km by 100 km zone stretching away from the Antarctic coast at 18°W off Riiser–Larsen Ice Shelf (Weddell Sea). They estimated a yearly passage of more than 950 icebergs for this area. The same work reported that the iceberg concentrations in a 300 km by 100 km strip east of the Antarctic Peninsula near to the Larsen Ice Shelf was lower than that for the Weddell Sea (yearly passage of
~150 icebergs). However, the near coast concentration of icebergs was 2 times higher than for the open ocean. A similar picture is drawn by Young et al. (1998) for the east Antarctic. They reported the width of the strip where icebergs concentrate to be 140-160 km and only sometimes up to 550 km off the coastline.
The higher near shore concentrations are caused by the Coriolis-force that results in a southward directed impulse, pushing icebergs towards the coast (Gladstone et al.
2001). Notably, the magnitude of this force is proportional to the size of the iceberg.
Thus smaller icebergs can easier leave the shelf zone, whereas bigger icebergs are trapped within the east wind drift.
On their way along the coast, icebergs do influence the ecosystem in a number of ways. Only giant icebergs may influence the pelagic ecosystem directly. However, the degree of disturbance is then very high. Examples are fragments of the giant B- 15 iceberg (~10,000 km²), which calved on March 2000 off the Ross Ice Shelf, which hindered the normal sea ice drift in the southwestern Ross Sea. This resulted in a much higher sea ice cover than normal and reduced the local primary production up to 40% (Arrigo et al. 2002). Similar, the C-19 iceberg, calved 26 months later in the same region, reduced the primary production as much as 90% (Arrigo and van Dijken
Error! Style not defined.
12
However, such giant icebergs are rare events. The influence of smaller icebergs on the pelagic ecosystem is assumed to be low (Schodlok et al. 2005). The main phenomena may arise from freshwater influx by meltwater, affecting the oceanographic parameters in the vicinity of the iceberg.
Principle mechanisms of iceberg-seafloor interaction in the Antarctic
Figure 6, Icebergs alter the flow pattern around them. Thus they have an indirect impact on the sea floor even if no physical impact occurs, e.g. accelerated flows causing resuspension of sediments.
Figure 7, Direct impact on the sea floor. Icebergs touch the seafloor and slip over it. This causes severe damage to all organisms, leaving devasted areas with characteristically plough marks. In front of the moving block, a rampart of sediment is heaped up. The sediment is turned over and partially resuspended.
Figure 8, If an iceberg has scoured and moved into too shallow water it gets trapped.
Changing currents, e.g. tidal movements, may cause the iceberg to seesaw, causing a pump-mechanism at the grounding zone. This may cause substantial resuspension and creates local debris slides on the sides (redrawn from Lien et al. 1989).
Error! Style not defined.
14
1.9 Influence of icebergs on the Antarctic sea floor
The influence of icebergs on the sea floor and benthic communities can be manifold.
Most impressive is the physical impact on the sea floor (see Figure 6,Figure 7 and Figure 8). The high mass and impulse of an iceberg lead to easily observable scour or plough marks at the impact zone. In general, iceberg plough marks resemble a flattened U with some small ridges on the sides. Like a bulldozer, the icebergs push the sediment in front of it into a berm. Numerical modelling indicates that this movement affects the sediment up to tree times the scour depth (Yang and Poorooshasb 1997). While moving, this front berm is turned over and the sediment is partially resuspended. Parts of the sediment are bulldozed to the sides and piled up into ridges along the iceberg track. The height of these berm ridges varies from few centimetres up to several metres. Rearic et al. (1990) estimated that in the shallow Harrison Bay (Alaska), coarse-grained material (>63 µm) is moved as far as 7 m in the direction of the ice movement. Finer sediments (<63 µm) can be transported even more than 500 m due to bottom currents.
When a iceberg is nearly of finally stranded it may seesaw due to changing currents, e.g. tidal movement. This results in strong currents around the contact zones (Lien et al. 1989). These currents can be the source of massive resuspension of sediments.
Together with the normal plough marks that can be several kilometres long and hundreds of meters wide (Lien et al. 1989), this “iceberg pump” mechanism could be a major source of sediment resuspension.
Based on the analysis of underwater video transects by Gutt and Starmans (2001) Potthoff et al. (2006) estimated 25 - 125 grounding events per year for a 300 km² region of the Weddell Sea, depending on water depth and topography. This leads to a rough estimate of about 1 - 5 grounding events per 10 km² for the whole shelf (0- 500 m water depth). Assuming a disturbance interval of roughly 250 - 350 years for each m² of the sea floor (numbers based on Gutt 2001), approximately 0.004- 0.003 % of the whole shelf (<500 m depth) is disturbed each year. Assuming the total shelf area (<500 m) to be between 1.2-2.2*106 km² (Barnes (1986), Gutt (2001); in comparison Clarke and Johnston (2003) list 2.97*106 km² <1000 m not under a permanent ice shelve) this results in a yearly perturbation area between 3.1-9.2*103 km².
2. Ecological Modelling
“The human intellect is impossible to think in other ways than in models. Also we experience nature always in a reduced way. No one has access to all aspects of an object in total and is able to store this information. To get essential information, we are forced to select and to abstract. Already when thinking what is essential, the formulation of a theory begins.”1 (Freely adapted from Wissel 1989)
Ecological modelling is, like statistics, a very valuable tool for ecologists. In general, a model is the simplified representation of a phenomenon from a certain point of view.
The reason to model or simulate natural phenomena is either to get a better understanding of the involved processes, to make predictions about the system response or simply to conduct experiments that cannot carried out in reality. To achieve such goals, several kinds of modelling approaches exist. The following passage references mainly to Wissel (1989), who coarsely classified three different model types used in the field of ecology:
• Conceptional models
• Descriptive models
• Simulation models
2.1 Conceptional Models
Conceptional models are used to create a better theoretical understanding of processes and phenomena (Wissel 1989). They are closest to classic mathematical models. This class of models often implies strong abstraction and reduction. An example for this model type is the well-known Lotka-Volterra model of two interacting species. The absolute focus of this model is the interaction of two species. External sources of mortality other than inter- and intraspecific competition of the two species are ignored.
1 “Der menschliche Geist ist unfähig, anders als in Modellen zu denken. Wir machen uns auch von der Natur immer vereinfachte Bilder. Kein Mensch kann alle Eigenschaften eines Objektes erfassen und alle erreichbaren Informationen darüber abspeichern. Er ist also gezwungen auszusondern, zu abstrahieren, um wesentliche
Error! Style not defined.
16
No one will ever assume that any natural population can be represented by such a simplified set of parameters. However, the Lotka-Volterra model has become one of the most influential models in theoretical and practical ecology even due to its simplicity.
It is worth to mention that the empirical background of some conceptional models is sometimes vague. For example, the well-known lynx-hare data cited in most text books (e.g. Begon et al. 1998). This data set has widely been used to explain models of the Lotka-Volterra type. However, only few acknowledge the fact that the data for the hare population comes from Eastern Canada, while the lynx data from Western Canada (see discussion in Hall 1988). Thus it is unlikely that both populations did directly interact with each other. Therefore, theoretical ecology and the search for general and universal ecological laws have often been criticized (Hall 1988, O'Hara 2005).
2.2 Descriptive Models
In contrast to conceptional models, descriptive models are used to classify and characterise systems. The goal of such models is to summarise and condense all available information of a system. Therefore, such characterisations may be used to extrapolate the system behaviour in the first hand and not to gain a mechanistic understanding. A very basic descriptive model is thus a simple linear regression.
Even an ecological index like Shannon’s H may be seen as simple descriptive model of the diversity of a community.
2.3 Simulation Models
The last class, simulation models, are in between both other types. Simulations are often created to obtain knowledge on the system behaviour and involved processes.
Predicting the future state of a system, e.g. the population size of some commercial exploited stocks under different harvesting regimes, is also a common modelling task.
Simulation models are typically applied when the system is rather complex and a simple, mathematical solution is impossible. But also even quite simple systems may require simulation models to understand their behaviour. For example, cellular automats, like Conway’s Game of Live or the Travelling Ant, have quite simple rules
but the final spatial patterns they create are not predictable and highly complex (see box “The Travelling Ant Example”).
In general, a perfect model should be realistic, precise and universal. According to the principle of Occam’s razor2 the model should further not contain unnecessary information and processes. Creating simulation models is thus the art of finding the necessary level of abstraction and complexity. Incorporating more and more details into a model, the covered processes become more and more realistic. On the other hand the uncertainty about the parameterisation rises. Thus, an optimal level for complexity exists (Wissel 1989).
However, models can just represent factors and mechanisms that were feed into the model a-priori. In principle this holds true even for models using evolutional algorithms. Therefore, results are valid only under the assumptions and restrictions defined also a-priori. It is important to be aware of this. Models are just a tool helping to extend our intellect. Caswell (1988) states that models are to theoretical problems what experiments are to empirical problems. They can be used for checking if hypotheses can work, to find logical cues of concepts and to stimulate a further discussion on the ideas. But they must never be used without critically evaluating their results.
2 “Entia non sunt multiplicanda praeter necessitatem” (Entities should not be multiplied beyond necessity), attributed to William Ockham (1295–1349)
“Occam's razor states that the explanation of any phenomenon should make as few assumptions as possible, neliminating, or "shaving off," those that make no difference in the observable predictions of the explanatory hypothesis or theory. In short, when given two equally valid explanations for a phenomenon, one should embrace the less complicated formulation.
Furthermore, when multiple competing theories have equal predictive powers, the principle recommends selecting those that introduce the fewest assumptions and postulate the fewest hypothetical entities. It is in this sense that Occam's razor is usually understood” (Occam's razor. (2006, September 28). In Wikipedia, The Free
Error! Style not defined.
18
The Travelling Ant Example
The Travelling Ant is a cellular automat like Conway’s Game of life. The ant lives in a gridded, infinite plane. Each grid element can be black or white. If the ant is on a white grid cell, it makes the cell to become black, turns left and moves one step forward. On a black cell, it changes the colour to white, turns right and then moves forward. These two simple rules can create astonishing complex patterns. On a complete, initially white plane, the ant starts its journey and by time a very regular, ladder alike pattern emerges. However, if the plane is randomly initialised with black and white cells, it is impossible to predict when and where this characteristic ladder shows up. Although the behaviour of the ant is completely determined and the starting conditions are fully known, you have to run through the complete simulation to answer this question.
The ant starts in the middle of the picture After 8000 steps a complex, however unordered movement pattern can be observed.
After about 10000 steps the ant starts to build the characteristic ladder like ant street.
On a randomly scattered plane it is impossible to predict when and where the ant street will be build
2.3.1 Spatial explicit models
Many models do not explicitly contain spatial information. Many conceptional models e.g. assume that every entity can interact freely with each other, the so called mean- field assumption. Also for models targeting at population sizes, spatial information is possibly unnecessary and should be left out (Occam’s razor cut’s here). However, as Murrell et al. (2001) pointed out, the spatial arrangement has a great influence on a system and every natural system is subject to spatial phenomena.
Spatial explicit models contain properties describing the spatial relation of their entities, e.g. their position in space or distance to each other. The addition of a further dimension “space” to a model can produce completely different results.
Coexistence in classical (“un-spatial”) models for example is only possible under some restrict assumptions. Including special aspects, coexistence can simply emerge due to the (spatial) separation of the species.
2.3.2 Individual Based Models
Individual based models (IBM) or agent based models (ABM) are a special class of simulation models. Basic elements of such simulations are one or several entities, typically representing a single individual and rules or directives that describe their behaviour. The meaning of individual may even cover a group of identical individuals e.g. a fish swarm. In this case modellers often refer to superindividuals. Despite a simple set of parameters and behaviour rules describing each entity the emerging complexity of individual based models, the final interplay of entities among themselves and their environment, can be very high.
IBMs offer a simple and easy way to describe complex, spatial systems. A further advantage is the fact of an easy and intuitive access of non-modellers to the underlying ideas. However, the exact formal description of IBMs can be difficult (Grimm et al. 2006).
2.3.3 Simulation Models and Biodiversity
Biodiversity, or biological diversity, has been defined as:
Error! Style not defined.
20
“Biological diversity” means the variability among living organisms from all sources including, inter alia, terrestrial, marine and other aquatic ecosystems and the ecological complexes of which they are part; this includes diversity within species, between species and of ecosystems. (DIVERSITY 1992)
This implies that diversity comprises the abundance of species and their spatial layout. Spatial explicit IBMs include both the abundance and spatial arrangement of individuals. Therefore, spatially explicit IBMs are especially suited to study biodiversity in a simulated habitat and the factors influencing it.
2.4 The Modelling Cycle
Modelling or creating computer simulations involves several steps:
1) Formulation of the aim or question to be solved 2) Identification of the key processes and mechanisms 3) Formulation of a conceptional model
4) Creating an appropriated model representation (i.e. a computer programme) 5) Verification if the model performs according to the conceptional model 6) Validation if the model reproduces reasonable results
7) Conducting the actual experiments 8) Interpretation of the results
Especially verification and validation are very important steps. If the model does not satisfy these points, the interpretation of the results will be, at best, questionable. If the verification fails (the model produced no suitable results, i.e. in a marine ecosystem the filter feeders prey on the whales), the model representation may be wrong. If the validation fails (i.e. the filter feeders prey on plankton but they never starve if they fail to catch something) there are strong hints that the model does not include all necessary process. Therefore, the above listed steps may be repeatedly carried out in different orders during the modelling process.
3. The Aim of this study
Due to their remoteness, marine systems in general and the Antarctic shelf in special are difficult to investigate and data on species traits like dispersal properties or life spans are spare. It is also unclear, how specific species traits and the environment influence the coexistence and diversity in these systems. The aim of this study is therefore to investigate the influence of disturbances and selected species traits on the diversity of marine systems and possible effects of the special Antarctic environment using computer simulation models.
The Antarctic shelf harbours a wide range of highly adopted species and is subject to intense disturbance events caused by grounding icebergs. Such grounding events cause catastrophic disturbance on the sea floor and are lethal to most affected individuals. After a disturbance a succession occurs, from diverse, unpredictable pioneer assemblages to a more defined climax state (Gutt and Piepenburg 2003, Teixido et al. 2004). This succession is influenced by the dispersal potential of the colonisers. In general, many marine sedentary species disperse via larvae. The larvae release often happens very seasonal, i.e. occurs over a short time period. If the larvae respond similar to their environment, they stay together and become dispersed as a group or swarm. This will result in clumped dispersal and settlement pattern. What is the outcome of such dispersal pattern? How do the dispersal traits of Antarctic species interact with such pattern? In general the dispersal potential of most Antarctic species is considered to be low. Theoretical work based on terrestrial systems has shown that short dispersal can foster competitive displacement (Bolker and Pacala 1999). In the first manuscript I focus in on the following questions:
• How does clumped dispersal influence the coexistence and diversity?
• How does it interact with the proposed short dispersal of most Antarctic species?
• Are there differences in the performance of terrestrial and aquatic systems?
The second manuscript targets the general coexistence of two species competing for space. Coexistence in this case refers to the long-term local persistence of both species. It is known that such coexistence can be mediated by a competition-
Error! Style not defined.
22
pioneer, has to explore new habitats before the superior species arrives and displaces the pioneer. If new habitats are created by disturbances (like the scour marks on the Antarctic shelf) the disturbance regime (disturbance size and frequency) plays an important role. It is reasonable to assume that dispersal implies cost for a species. Therefore dispersal should be effectively and cost efficient. The second manuscript studies the following questions:
• Is there a dispersal distance threshold, a minimum dispersal distance that allows the species to persist under the aspect of cost efficiency?
• How does this threshold depend on the disturbance regime?
• How does the proposed prolonged longevity of Antarctic species affect the dispersal distance?
Besides biological factors the succession after a disturbance is controlled by several abiotic factors. The disturbance regime, namely the average size of a single disturbance event, the frequency of such events and their severity interact with the species traits and influence the succession process. The last manuscript analyses the role of these disturbance properties for the succession process. It uses a simulation of a multispecies community on the Antarctic shelf, which is subject to intense disturbance. With this simulation model the questions are explored:
• In the same time interval numerous small disturbances can perturb the same total area as rare large disturbances can. Is only the total area important or lead numerous small disturbances to other results as infrequent but larger disturbances?
• Which role has the average size of a single disturbance?
• How important is the disturbance frequency?
• How do possible survivors influence the recovery process?
The following sections contain summaries of the publications and manuscripts describing the scientific work carried out during this thesis. The first two manuscripts have been already published; the third is ready to be submitted. After the summary of the manuscripts a brief discussion of the results follows. Detailed model descriptions, results and discussions can be found in the text of the manuscripts in the chapters 7- 9.
As a last section a detailed manual for SIMBAA (Simulation Model of Benthic Antarctic Assemblages) follows. The development of SIMBAA was a main task during this thesis. The source code of SIMBAA spans about 35 single files with together more than 16.000 lines of code and was completely written from the scratch, using Borland Delphi 7.0 Professional. A portable pseudo-random-generator was taken form the Numerical Recipes Series (Press et al. 1991). Additionally, a freeware colour space conversion routine written by Grahame Marsh and some code lines to display the compile date of the executable programme were taken from the Internet (see SIMBAA manual for details). This foreign code is not essential for the model, but primary reduced the development time. In case of the pseudo-random-generator the selection of a fully accessible and proven code simplifies the validation, quality and reusability of the whole model.
Error! Style not defined.
24
4. Publications and Manuscripts
4.1 “Clumped Dispersal and Species Coexistence”
The idea, model programming, evaluation of the result and the first draft of the manuscript was done by myself. The co-authors contributed to the writing and discussion of the final manuscript.
4.1.1 Bibliographic record
Title: Clumped Dispersal and Species Coexistence
Authors: Michael Potthoff, Karin Johst, Julian Gutt, Christian Wissel
Status: published, Ecological Modelling Volume 198 (1-2), Pages 247-254 doi:10.1016/j.ecolmodel.2006.04.003
4.1.2 Short summary of ideas, problems, solutions and results:
Many sedentary marine species disperse their propagules passively. Compared to air, the high water density allows species to easily adopt the buoyancy of their larvae to be negative, positive or neutral. Thus dispersal using oceanic currents can involve comparable few metabolic costs. However, the risks of becoming dispersed to unsuitable habitats or falling prey during the dispersal phase raises as longer the dispersal lasts. Often planktonic larvae exhibit sophisticated vertical migration patterns, triggered by intrinsic and/or external factors, to avoid these risks.
Larvae are often released only over a short period, e.g. within a few days during the mass spawning observed at some coral reefs. It must be assumed that factors triggering larvae migration pattern are species specific. Thus larvae of one species respond similar to their environment. This may lead to a separation and concentration of species specific larvae in specific water masses. Consequently the arrival of such larvae swarms and the settlement co-varies. On a local scale this will resemble a wave or pulse like larvae and settlement pattern. On a coarser spatial resolution a patchy pattern will emerge.
To investigate the role of such clumped and patchy dispersal and settlement of larvae for coexistence and diversity we implemented a spatial explicit IBM. The species in
the model differed only in their dispersal traits, all other traits where equal. The key idea was to introduce a two-phase dispersal mechanism: in the first phase the average dispersal distance of the larvae swarm is determined. The swarm virtually centres around the resulting point. Swarm members are then distributed around this point. This dispersal mechanism is very flexible. Depending on the chosen values for the dispersal distances either patchy and clumped or an isotropic, source centred larvae pattern can be generated.
We analysed the role of clumped dispersal for diversity (in sense of species richness) and its competitive performance with isotropic dispersal with this model. In all tested cases clumped dispersal allowed multi-species coexistence over long time periods, whereas isotropic dispersal fostered competitive displacement and quickly lead to the extinction of most species when only one dispersal strategy (clumped/isotropic) was available. When both dispersal strategies competed, coexistence was possible when clumped dispersal had the superior dispersal distance (depending on the environment).
4.2 “How to survive as a pioneer species in the Antarctic benthos: minimum dispersal distance as a function of lifetime and disturbance”
The idea, model programming, evaluation of the result and the first draft of the manuscript was done by myself. The co-authors contributed to the writing and discussion of the final manuscript.
4.2.1 Bibliographic record
Title: How to survive as a pioneer species in the Antarctic benthos:
minimum dispersal distance as a function of lifetime and disturbance
Authors: Michael Potthoff, Karin Johst, Julian Gutt Status: published, Polar Biology 2006, 29 543-551
doi: 10.107/s00300-005-0086-1
4.2.2 Short summary of ideas, problems, solution and results:
Under the assumption that dispersal induces costs for a population, species should disperse cost efficient, i.e. only as far as the colonisation rate of new habitats allows
Error! Style not defined.
26
the persistence of the population. However, if individuals live for more than one season, they have several chances for successful dispersal. Thus the dispersal distance should be a function of habitat distance and lifespan. In a dynamic environment, where new habitat is created by disturbances, the disturbance regime plays also an important role.
The aim of this study was to find out if a threshold in the dispersal distance exists that allows a species to persist in a dynamic environment and how this threshold depends on species longevity and the disturbance regime. We implemented a spatial explicit IBM with two species living in a dynamic environment. This environment is characterised by catastrophic disturbance events. Effected regions are completely disturbed and become free of any inhabitants. One of the modelled species depends on the colonisation of such free areas as it can only recruit in the absence of the other species. Thus it is the inferior coloniser and represents a classic pioneer species. The second, superior species is able to competitively displace the pioneer and quickly colonise all space not utilised by it.
It is well known that coexistence of both species is possible under a dispersal- colonisation trade-off (Tilman 1994). This means that the inferior species must have a dispersal distance that allows it to colonise new habitats before the superior species arrives. However, the exact threshold and the dependency of the dispersal distance on the disturbance regime and lifespan of the pioneer are unclear.
The key idea of our implementation was to use circular disturbance events and circular dispersal shadows. This allowed us to easily compute distances between different patches using a simple Euclidian distance. Experimental manipulation of the disturbance regime, dispersal distance and life span allowed us to determine a dispersal distance threshold for the persistence of the inferior species.
The results show that the dispersal distance must be at least slightly higher than the average distance to the next free habitat. Thus, with raising disturbance intensity, more disturbances create more free space in the vicinity of a pioneer population.
Consequently, the dispersal distance sufficient to allow the pioneer to persist can be reduced. However, under a very high disturbance regime the dispersal distance must raise again as the probability for a catastrophic disturbance rises. Thus the threshold shows a U-shaped relationship to the disturbance intensity.
Long living species can cope with low dispersal distances as they have more chances for successful dispersal. Virtually a doubled lifetime doubles the chances,
e.g. it creates a lower average new habitat distance. On the other hand the probability of a catastrophe rises. The interplay of both processes brings about a non- linear, hyperbolic relationship of dispersal distance and lifespan.
Our model indicates that a species could persist with half the dispersal distance if it could raise its lifespan three to four times. Thus, in a dangerous world with high dispersal costs, becoming older is a good strategy. Caused by the hyperbolic relationship, especially short living species can profit much if they can extend their lifespan.
4.3 “How the disturbance severity drives the benthic diversity on the Antarctic shelf”
The idea, model programming, evaluation of the result and the first draft of the manuscript was done by myself. The co-authors contributed to the writing and discussion of the final manuscript.
4.3.1 Bibliographic record
Title: How the disturbance severity drives the benthic diversity on the Antarctic shelf
Authors: Michael Potthoff, Karin Johst, Julian Gutt Status: unpublished
4.3.2 Short summary of ideas, problems, solution and results:
Generally disturbance is defined as a relatively distinct event in space and time that disrupts the ecosystem, community or population structure and changes resources, substrate availability or the physical environment (White and Jentsch 2001). On the Antarctic shelf physical disturbance by grounding icebergs is the major disturbance agent for the benthos, disrupting local communities and populations and influencing the local physical environment and substrate characteristics. The aim of this manuscript was to characterise the importance of such disturbances for the succession and diversity. In particular we were interested in how the disturbance size and frequency influenced the succession and recovery speed. Changes in flow and sediment conditions and their influence on the communities were not considered.
